A farmer’s job is quite hard. Good results depend on several factors such as the type of Soil, Water, Fertilisers, Pesticides. Excessive use of chemicals can damage a crop and can also cause harm to its consumers. To top it off, global warming has created unpredictable weather patterns and has the potential to destroy entire seasons of crops without much notice.

To solve this we will build out a system that will be able to predict the outcome of a crop season based on the various information we will collect. This system will be able to guide the farmer to use the right amount of water, fertilizer, pesticides. Predict the correct intervals to use these. Be able to detect important threats such as unexpected rodents and suggest corrective measures. The system will also be keeping account of the changing weather patterns and suggest deviations accordingly.

The system will use sensors and cameras to collect the following information from the field for real time prediction:
# Soil Properties
# Localized Weather Properties
# Collect images of the plantation and nearby areas
# Images of chemicals use, if possible quantity of chemicals used by sensors
# Water properties
# Macro weather condition

We will build out the model using the data provided by the Ministry of Agriculture and Farmer’s Welfare, Government of India and various other open data sets. The model itself will be using KNN Algorithm. It is widely used in text categorization, predictive analysis, data mining and image recognition and will be suitable for our use case.

We will use KNN algorithm on FPGA based heterogeneous computing systems using OpenCL. Based on FPGA's parallel pipeline structure. Use of FPGA will improve the efficiency of the solution compared to a conventional GPU based KNN algorithm implementation.

Project Proposal

1. High-level project introduction and performance expectation

Purpose of Design

Agriculture is the primary source of livelihood for more than 58% of people in India. Agriculture done in most parts of the country still is largely a manual process. These processes are generally passed from one generation to another and often things get lost in translation. 

A farmer’s job is quite hard. Good results depend on several factors such as the type of Soil, Water, Fertilizers, Pesticides. Excessive use of chemicals can damage a crop and can also cause harm to its consumers. To top it off, global warming has created unpredictable weather patterns and has the potential to destroy entire seasons of crops without much notice.

To solve this we will build out a system that will be able to predict the outcome of a crop season based on the various information we will collect. This system will be able to guide the farmer to use the right amount of water, fertilizer, pesticides. Predict the correct intervals to use these. Be able to detect important threats such as unexpected rodents and suggest corrective measures. The system will also be keeping account of the changing weather patterns and suggest deviations accordingly.

Application Scope

The system will use sensors and cameras to collect the following information from the field for real time prediction:

• Soil Properties

• Localized Weather Properties

• Collect images of the plantation and nearby areas

• Images of chemicals use, if possible quantity of chemicals used by sensors

• Water properties

• Macro weather condition

We will build out the model using the data provided by the Ministry of Agriculture and Farmers Welfare, Government of India and various other open data sets. The model itself will be using KNN Algorithm. It is widely used in text categorization, predictive analysis, data mining and image recognition and will be suitable for our use case. 

Targeted Users

We will target the users with following profile:

• Small and Medium size farmers in India. 

• Deployment area has Jio cellular connectivity

• Farmer should be able to operate a smartphone

Why Intel Cyclone 5 FPGA

We will use KNN algorithms on FPGA based heterogeneous computing systems using OpenCL. Based on FPGA's parallel pipeline structure. Use of FPGA will improve the efficiency of the solution compared to a conventional GPU based KNN algorithm implementation.

We are using Intel Cyclone 5 FPGA because it provides the following advantages:

• First party support for OpenCL directly from Intel

• Can be easily integrated with Linux using the DE10 Nano kit. This allows us to use the traditional SoC for common things while using FPGA for OpenCL code and parallelization.

• Cost effective while delivering the performance of 110 LEs

2. Block Diagram

3. Expected sustainability results, projected resource savings

This project addresses sustainability in the following ways:

• It will save water used for irrigation by carefully predicting and providing the exact water used in harvesting the crops. Estimated reduction in water wastage is around 30%.

• It will reduce the harm to soil by carefully managing the amount of fertilizers used in the soil. This will help keep the fertility of the soil higher by 20% for the next crop season. 

• Increase the output using the existing resources by 15%.

4. Design Introduction

5. Functional description and implementation

6. Performance metrics, performance to expectation

7. Sustainability results, resource savings achieved

8. Conclusion